AI should not leave women behind

Last year was a landmark year for gender equality, marking the 30th anniversary of the Beijing Declaration and Platform for Action — a historic global commitment to gender equality. Recognizing this, global leaders and experts assembled in 2025 on the margins of the 69th session of the Commission on the Status of Women, to launch the Beijing+30 Action Agenda, laying out six key actions to accelerate empowerment of women and girls.

One of these actions is "a digital revolution" emphasizing the importance of ensuring that women and girls have equal access to new skills and opportunities that technology is opening up, while also being able to shape how these innovations are designed.

As we enter 2026, this focus could not be more timely — particularly with the rapid evolution of artificial intelligence. Today, AI is transforming almost every aspect of our lives. Yet as this powerful technology races ahead, women are being left behind.

Globally, less than a quarter of AI professionals — and less than a third of STEM workers — are women. The gap is growing. In fact, a greater proportion of women were tech leaders in the 1980s than today. The result is stark: men make most of the decisions shaping tomorrow's tools. They design AI models, determine how they are tested, select training data, and decide how our data are used.

Ensuring more women become AI decision-makers is vital — for equality, and for humanity.

First, when men and women develop AI together, technology represents everyone more accurately. Without sufficient input from women, AI often gets it wrong for them — including on critical issues, like health. For example, Fitbit devices are reportedly less capable of recording accurate results for shorter users.

Second, evidence suggests technologies shaped by women are more accountable to all. Researchers have found that women, on average, consider a wider range of people when making decisions, and a wider range of risks, including misinformation.

Third, when women do develop AI-driven solutions, these often have a profound impact on the challenges they face. Consider 16-year-old Bohlale Mphahlele in South Africa, who invented an earring with a hidden security camera and GPS to help combat human trafficking and gender-based violence.

Clearly, women can strengthen AI for everyone. But they face unique and persistent barriers in getting there, which society must urgently address.

One risk is that women's underrepresentation in STEM becomes entrenched through AI itself. Recruitment algorithms, for example, learn from historical patterns. After a decade of receiving mostly male CVs, one e-commerce firm's hiring system began excluding otherwise qualified applications that included the word "woman". The system was ultimately disbanded. But AI-driven inequality will only be overturned more broadly if companies proactively programme against bias, by enabling new patterns, not reinforcing old ones.

Policy and workplace reforms are also essential. Policymakers and employers can help women commit to demanding, but fulfilling STEM careers by expanding flexibility. Hybrid work options, such as flexible remote work when needed, can be transformative for women, who still shoulder a disproportionate share of family responsibilities, including child and elderly care. Their domestic load can also be eased by expanding parental leave particularly for men, who have just days, in many cases. Only with greater equality at home can we achieve equality at work.

This is becoming even more urgent, as jobs predominantly held by women are twice as likely to be lost to AI than those held by men. Clerical and administrative roles, where many women are currently concentrated, are among the most vulnerable. Women and girls must therefore be trained and encouraged to grow with AI, rather than being replaced by it.

Finally, change must also come from the bottom up. Policymakers can incentivize academic institutions and employers to train and recruit more women in AI. Schools can boost early exposure by developing courses and activities that engage girls to explore AI.

UNDP is committed to closing gender gaps in STEM education and employment. In China, we partner with NGOs and UN Volunteers on our HER Digital Future bootcamp since 2023. So far, more than 5,000 girls across over 90 rural schools have learned about emerging technologies, and how to design digital solutions to address challenges in their local communities such as waste management.

But mindsets must also shift at home. How parents talk to daughters, the kind of books and activities they introduce, and the responsibilities they entrust to them can profoundly shape curiosity and confidence — key qualities for any future tech leader.

In fact, the world's first computer programmer was a woman, whose parents did just that. Ada Lovelace, encouraged by her parents to study mathematics, put forward an idea that changed the world: that numbers could represent something other than quantity. This insight sparked the revolution from calculation to computation. Every year in October, Ada Lovelace Day celebrates overlooked achievements of women in STEM, reminding us of the great potential that is still to be unlocked.

We need all the STEM leaders we can get. Because science and technology are critical to solving the planet's most complex and urgent challenges. Our window to avert the worst impacts of climate change is closing, with the 2030 deadline to meet the Sustainable Development Goals drawing near — including Goal 9 on innovation and Goal 5 on gender equality. Empowering women and girls in STEM would accelerate progress across all these, by ensuring that the technology that defines our future truly represents, and reaches, everyone.

The author is the Resident Representative of the UNDP in China.

The views don't necessarily represent those of China Daily.

If you have a specific expertise, or would like to share your thought about our stories, then send us your writings at opinion@chinadaily.com.cn, and comment@chinadaily.com.cn.